Although medication and supportive care have been effective in reducing the duration of hospital stay of schizophrenics, relapse continues to be a major problem. British studies suggest that the best single predictor of outcome is the level of emotion expressed by family members when individually discussing their relationship with the schizophrenic patient during a standardized interview. (Camberwell Family Interview, CFI.) Where high levels of expressed emotion (EE) are present, continued antipsychotic medication plus reduced contact between the patient and the family member appear necessarY to reduce the high risk of relapse. Where low levels of expressed emotion are present, the risk of relapse is low regardless of the services provided. The objectives of the proposed research are to provide specific conjoint family interventions to reduce expressed emotion and, thereby, decrease the probability of relapse in high EE families. Forty (40) patients, diagnosed schizophrenic with the Present State Examination, who are living in households containing one or more high EE family members, will participate in the evaluation of the comparative effects of family therapy and individual supportive therapy. All patients will receive optimal doses of chlorpromazine. A multi-level battery of outcome measures will be administered before and after intensive treatment and during the two years after the start of treatment. Measures will include assessment of psychopathology, social functioning, life events, coping skills, family relationships and attitudes, family communication and problem-solving skills, length of stay in the community, plasma levels of medication, and cost effectiveness.